Combining Images When a Face Is Present

1. An apparatus comprising: an image sensor that captures image data of a photographic subject matter; a processor; and a memory that stores code executable by the processor to: calculate a first exposure value for the photographic subject matter, wherein the first exposure value is calculated from the photographic subject matter as a whole; determine whether a face is present within the photographic subject matter; calculate a second exposure value in response to determining that a face is present within the photographic subject matter, wherein the second exposure value is calculated to expose the face; capture a first image using the first exposure value; capture a second image using the second exposure value; and create a third image from the first image and the second image, wherein creating the third image from the first image and the second image comprises the processor: determining whether a background region of the second image is overblown; compositing the first image and second image to form the third image in response to the background region of the second image being overblown; and discarding the first image and saving the second image as the third image in response to the background region of the second image not being overblown.

2. The apparatus of claim 1 , wherein compositing the first image and second image to form the third image comprises the processor: identifying a face region within the second image; and compositing the face region into the first image to form the third image.

3. The apparatus of claim 2 , wherein compositing the first image and second image to form the third image further comprises the processor: computing facial feature points in the face region; computing facial feature points in a corresponding region in the first image, the corresponding region comprising the face; and transforming the face region based on the computed facial feature points, wherein compositing the face region into the first image to form the third image occurs in response to transforming the face region.

4. The apparatus of claim 3 , wherein transforming the face region based on the computed facial feature points comprises the processor: generating a transformation matrix using the facial feature points in the face region and the facial feature points in the corresponding region.

5. The apparatus of claim 1 , wherein the processor further: calculates a reprojection error from a face region in the second image and a corresponding region in the first image, the corresponding region comprising the face; and compares the reprojection error to a threshold error amount, wherein creating the third image from the first image and the second image further comprises the processor: merging the first image and second image to form the third image in response to the reprojection error being less than the threshold error amount; and discarding the first image and saving the second image as the third image in response to the being greater than or equal to the threshold error amount.

6. The apparatus of claim 5 , wherein merging the first image and second image to form the third image comprises the processor: transforming the face region using a transformation matrix; replacing the corresponding region in the first image with the transformed face region; and blending the transformed face region with the first image.

7. The apparatus of claim 1 , wherein compositing the first image and second image to form the third image comprises the processor blending only a face region of the first image and second image.

8. The apparatus of claim 1 , wherein the processor further determines whether the apparatus moves more than a threshold amount in a time between the capture of the first image and the capture of the second image, wherein creating the third image from the first image and the second image further comprises the processor: merging the first image and second image to form the third image in response to the apparatus not moving more than the threshold amount; and discarding the second image and saving the first image as the third image in response to the apparatus moving more than the threshold amount.

9. The apparatus of claim 1 , wherein the first exposure value is calculated to correctly expose a background.

10. A method comprising: calculating, by use of a processor, a first exposure level for a photographic subject matter; determining whether a face is present within the photographic subject matter; calculating a second exposure level in response to determining that a face is present within the photographic subject matter; capturing a first image using the first exposure level; capturing a second image using the second exposure level, the second exposure level calculated to expose the face; and creating a third image from the first image and second image, wherein creating the third image from the first image and the second image comprises: determining whether a background region of the second image is overblown; compositing the first image and second image to form the third image in response to the background region of the second image being overblown; and discarding the first image and saving the second image as the third image in response to the background region of the second image not being overblown.

11. The method of claim 10 , wherein compositing the first image and second image to form the third image comprises: identifying a face region within the second image; identifying a corresponding region in the first image, the corresponding region comprising the face; transforming the face region using a transformation matrix based on the face region and the corresponding region; and blending the corresponding region in the first image with the transformed face region.

12. The method of claim 11 , wherein transforming the face region using a transformation matrix comprises: computing facial feature points in the face region; computing corresponding facial feature points in the corresponding region; and generating the transformation matrix using the facial feature points and the corresponding facial feature points, wherein the method further comprises blending the transformed face region with the first image.

13. The method of claim 10 , further comprising: calculating a reprojection error from a face region in the second image and a corresponding region in the first image, the corresponding region comprising the face; and comparing the reprojection error to a threshold error amount, wherein creating the third image from the first image and the second image further comprises: merging the face region into the first image to form the third image in response to the reprojection error being less than the threshold error amount; and discarding the first image and saving the second image as the third image in response to the being greater than or equal to the threshold error amount.

14. The method of claim 13 , wherein merging the face region into the first image to form the third image comprises blending only the face region with the first image, wherein the first exposure level is calculated to correctly expose a background.

15. The method of claim 10 , wherein compositing the first image and second image to form the third image comprises: merging only a face region of the first image and second image to form the third image in response to the background region of the second image being overblown.

16. The method of claim 10 , wherein creating the third image from the first image and the second image further comprises: determining whether a camera device capturing the first image moves more than a threshold amount in a time between the capture of the first image and the capture of the second image; merging only a face region of the first image and second image to form the third image in response to the camera device not moving more than the threshold amount; and discarding the second image and saving the first image as the third image in response to the camera device moving more than the threshold amount.

17. A program product comprising a non-transitory computer readable storage medium that stores code executable by a processor, the executable code comprising code to perform: calculating a first exposure value for a photographic subject matter; determining whether a face is present within the photographic subject matter; calculating a second exposure value in response to determining that a face is present within the photographic subject matter; capturing a first image using the first exposure value; capturing a second image using the second exposure value, the second exposure value calculated to expose the face; and selectively combining the first image and second image to form a third image, wherein selectively combining the first image and the second image comprises: determining whether a background region of the second image is overblown; compositing a face region of the second image into the first image to form the third image, in response to the background region of the second image being overblown; and discarding the first image and saving the second image as the third image in response to the background region of the second image not being overblown.

18. The program product of claim 17 , wherein selectively combining the first image and the second image further comprises: determining whether a camera device capturing the first image moves more than a threshold amount in a time between the capture of the first image and the capture of the second image; compositing a face region of the second image into the first image to form the third image, in response to the camera device not moving more than the threshold amount; and discarding the second image and saving the first image as the third image in response to the camera device moving more than the threshold amount.